Modeling method for driving performance prediction model of electric vehicle coasting brake energy recovery

Abstract

This invention relates to the field of electric vehicle testing technology and discloses a drivability prediction model and modeling method for electric vehicle coasting braking energy recovery system, comprising the following steps: Step 1, analyzing the coasting braking process of the electric vehicle coasting energy recovery system and extracting influencing factors and subjective evaluation sample data; the influencing factors are characteristic indicators related to the coasting deceleration characteristics of the electric vehicle coasting energy recovery system; Step 2, quantifying the influencing factors and screening them in conjunction with the subjective evaluation sample data; Step 3, establishing a subjective evaluation prediction model and using the screened influencing factors as explanatory variables of the subjective evaluation model; optimizing the subjective evaluation prediction model using each explanatory variable, and obtaining the final subjective evaluation prediction model. This invention can comprehensively and accurately evaluate and characterize the drivability of the coasting braking energy recovery system, and the evaluation consistency is high.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle testing technology, specifically to a drivability prediction model and modeling method for electric vehicle coasting braking energy recovery. Background Technology

[0002] The efficient use of energy by electric vehicles is key to leveraging their energy-saving and environmental advantages, and is the most effective way and method to solve environmental pollution and energy shortage problems. Since the birth of electric vehicles, their range performance has also been a major focus. In order to utilize this precious energy to increase the driving range, electric vehicles have designed a system that can store the kinetic energy of the car's movement in a certain way and convert it into power for the vehicle when needed—the coasting energy recovery system.

[0003] Specifically, electric vehicle coasting energy recovery systems achieve braking deceleration and energy regeneration through motor feedback torque, offering inherent advantages in braking efficiency and economy. However, due to the involvement of the motor drive system during braking, their drivability evaluation differs from that of traditional gasoline vehicles. Furthermore, the factors influencing drivability evaluation during braking deceleration are more complex. Current technologies typically rely on driver experience feedback regarding operational feel and driving comfort, or use objective parameters such as braking distance, braking intensity, and braking time to characterize braking performance. However, due to individual driving habits and subjective biases, the consistency of experience feedback is low. Objective parameter evaluations lack subjective feedback, have limited evaluation dimensions, and suffer from indicator limitations, failing to fully characterize drivability. Summary of the Invention

[0004] The present invention aims to provide a drivability prediction model and modeling method for electric vehicle coasting braking energy recovery, which can comprehensively and accurately evaluate and characterize the drivability of coasting braking energy recovery system, and the evaluation consistency is high.

[0005] To achieve the above objectives, the basic solution provided by this invention is as follows:

[0006] Option 1

[0007] The modeling method for predicting the drivability of electric vehicles by recuperation during coasting braking includes the following steps:

[0008] Step 1: Analyze the coasting braking process of the electric vehicle coasting energy recovery system and extract influencing factors and subjective evaluation sample data; the influencing factors are characteristic indicators related to the coasting deceleration characteristics of the electric vehicle coasting energy recovery system.

[0009] Step 2: Quantify the impact factors and, in conjunction with subjective evaluation sample data, screen the impact factors.

[0010] Step 3: Establish a subjective evaluation prediction model and use the selected influencing factors as explanatory variables of the subjective evaluation prediction model; optimize the subjective evaluation prediction model using each explanatory variable and obtain the final subjective evaluation prediction model.

[0011] Option 2

[0012] The electric vehicle coasting braking energy recovery drivability prediction model is obtained using the modeling method described in Scheme 1; it includes a linearized model:

[0013] ;in, , … These are the maximum decelerations. The linear regression parameters and the rate of change of maximum deceleration The linear regression parameters and deceleration coasting distance The linear regression parameters and maximum deceleration Quadratic regression parameters, maximum rate of change of deceleration Quadratic regression parameters, deceleration coasting distance Quadratic regression parameters, steady-state proportion of rate of change of deceleration The regression parameters, This is a disturbance term.

[0014] The working principle and advantages of this invention are as follows:

[0015] This scheme proposes influencing factors affecting drivability during coasting braking and defines a quantitative calculation method for these factors. Correlation analysis is then performed on these quantified factors to identify those highly correlated with subjective evaluation results. Finally, a subjective drivability evaluation prediction model is established using nonlinear regression to predict drivability during electric vehicle energy recovery, achieving an explanation ratio of over 77% and demonstrating good model fit. Specifically, this scheme constructs a subjective evaluation prediction model based on characteristic indicators (influencing factors) related to the coasting deceleration characteristics of the electric vehicle's coasting energy recovery system, combined with subjective evaluation sample data. This model comprehensively considers both subjective and objective dimensions of data, enabling a comprehensive and accurate evaluation and characterization of the drivability of the coasting braking energy recovery system, with high evaluation consistency. Attached Figure Description

[0016] Figure 1 This is a schematic flowchart illustrating an embodiment of the electric vehicle coasting braking energy recovery drivability prediction model and modeling method of the present invention;

[0017] Figure 2This is a schematic diagram comparing the prediction model results with the average actual subjective score at different initial speeds in the maximum energy recovery intensity mode, as shown in the embodiment of the electric vehicle coasting braking energy recovery drivability prediction model and modeling method of the present invention.

[0018] Figure 3 This is a schematic diagram of the vehicle coasting braking process, which is an embodiment of the electric vehicle coasting braking energy recovery drivability prediction model and modeling method of the present invention. Detailed Implementation

[0019] The following detailed explanation illustrates the specific implementation methods:

[0020] The basic implementation examples are as follows: Figure 1 The following is a modeling method for predicting the drivability of an electric vehicle's coasting braking energy recovery model, including the following steps:

[0021] Step 1: Conduct deceleration coasting tests on electric vehicles equipped with coasting energy recovery systems, analyze the coasting braking process of the electric vehicle coasting energy recovery system, and extract influencing factors and subjective evaluation sample data; the influencing factors are characteristic indicators related to the coasting deceleration characteristics of the electric vehicle coasting energy recovery system.

[0022] The subjective evaluation sample data includes deceleration intensity evaluation data, deceleration response evaluation data, and deceleration impact evaluation data.

[0023] The deceleration intensity is described based on the driver's body posture. Generally, the greater the deceleration intensity, the more unstable the driver's posture. Excessive or insufficient deceleration intensity will reduce comfort or sense of security. Deceleration response is judged based on the speed of the system's braking response during deceleration, including whether the system quickly meets the braking demand and whether the entire braking process is rapid. It should be within a reasonable range; excessive or insufficient response will reduce safety. Deceleration impact is judged based on the degree of impact and dizziness experienced by the driver during braking, and it should have an optimal solution within a reasonable range.

[0024] The influencing factors include the maximum deceleration corresponding to the deceleration intensity evaluation data. Extreme deceleration Root mean square of deceleration The time required to reach the deceleration inflection point and the deceleration coasting distance corresponding to the deceleration response evaluation data; the maximum rate of change of deceleration corresponding to the deceleration impact evaluation data. Range of deceleration rate of change Root mean square of the rate of change of deceleration Steady-state proportion of the rate of change of deceleration .

[0025] Step 2: Quantify the impact factors and, in conjunction with the subjective evaluation sample data, screen the impact factors.

[0026] Specifically, when quantifying the impact factor, it is quantified according to the following formula:

[0027] Where j represents the impact intensity, in units of... ; 'a' represents vehicle acceleration, in units of V represents vehicle speed, in units of... t represents time, in seconds (s).

[0028] ; ;in, For the deceleration range, unit ; , These are the maximum and minimum deceleration values, in units of... ; For the range of the rate of change of deceleration, unit ; , These are the maximum and minimum values ​​of the rate of change of deceleration, in units of... .

[0029] ;in, For the root mean square of the deceleration, in units ; , These are the start and end times, in units. .

[0030] ;in, The percentage of the steady-state rate of change of deceleration is dimensionless. The moment when the rate of change of deceleration first reaches a steady state, unit ; For the steady-state end of the rate of change of deceleration, the unit In this embodiment, the criterion for determining the steady state of the rate of deceleration is the fluctuation range of the rate of deceleration change value within ±0.05m / Duration 1s. During coasting braking, the electric vehicle coasting energy recovery system should minimize the transient impact on the driver. The longer the deceleration rate takes to reach a steady state (the larger the proportion of steady state), the more consistent the deceleration change per unit time, which can reduce dizziness during coasting braking and improve driving comfort.

[0031] Deceleration driving distance The distance traveled by an electric vehicle's coasting energy recovery system from the initial deceleration state during coasting braking to the first steady state is measured. This indicator represents the braking efficiency of the electric motor and can indirectly reflect the energy conversion efficiency of the electric vehicle's coasting energy recovery system.

[0032] Time required to reach the deceleration inflection point ;in, The moment when the initial state of coasting braking begins in the electric vehicle coasting energy recovery system; This refers to the moment when the deceleration reaches its inflection point during coasting braking. This indicator characterizes the system's (referring to the electric vehicle's coasting energy recovery system) response speed to achieve the expected braking demand. The moments during coasting braking are shown in the attached figure. Figure 3 As shown, the time required to reach the maximum deceleration is .

[0033] When screening influencing factors, correlation analysis is used; the correlation analysis includes the following sub-steps:

[0034] S1: Determine the correlation between influencing factors and subjective evaluation sample data through subjective judgment. In this embodiment, it is determined that there is a non-linear relationship between several initially selected influencing factors and subjective evaluation sample data. For example, in the deceleration intensity test, excessive maximum deceleration will reduce comfort, while insufficient maximum deceleration will weaken the sense of security.

[0035] S2: Visualize the quantified impact factor data using graph correlation analysis and determine the correlation between the data.

[0036] S3: If the correlation in S2 is linear, then correlation analysis is performed directly; if the correlation in S2 is curvilinear, then linear regression is used to transform the nonlinear variable relationship into a linear variable relationship before performing linear correlation analysis. In this embodiment, the results of the correlation analysis for each influencing factor are shown in Table 1 below:

[0037] Table 1. Results of Correlation Analysis of Influence Factors

[0038]

[0039] The selection criteria for the explanatory variables include: selecting the correlation coefficient from the correlation analysis method. The influencing factors are used as explanatory variables. Under this screening condition, it is possible to accurately select influencing factors that are highly correlated with subjective evaluations and are more targeted at driving performance, which helps to ensure the consistency of evaluation.

[0040] Step 3: Establish a subjective evaluation prediction model and use the selected influencing factors as explanatory variables of the subjective evaluation prediction model; optimize the subjective evaluation prediction model using each explanatory variable and obtain the final subjective evaluation prediction model.

[0041] In this step, when establishing the subjective evaluation prediction model, the explanatory variables selected through the above correlation analysis method, combined with the relationship model obtained through graphical correlation analysis, form the following subjective evaluation prediction model:

[0042] ;in, , … These are the maximum decelerations. The linear regression parameters and the rate of change of maximum deceleration The linear regression parameters and deceleration coasting distance The linear regression parameters and maximum deceleration Quadratic regression parameters, maximum rate of change of deceleration Quadratic regression parameters, deceleration coasting distance Quadratic regression parameters, steady-state proportion of rate of change of deceleration The regression parameters, This is a disturbance term.

[0043] Furthermore, the original nonlinear relational model is linearized by weighting the sample set; let = , = , = , = Then, the linearized subjective evaluation prediction model is:

[0044] ;

[0045] Regression analysis was performed using the explanatory variable data that met the conditions in the deceleration and coasting test. The final regression coefficients were then used to derive the prediction model (i.e., the final subjective evaluation prediction model):

[0046] .

[0047] Step 4: Input the test data into the subjective evaluation prediction model and evaluate the subjective evaluation prediction model.

[0048] The following test example is used to evaluate the subjective evaluation prediction model to demonstrate the effectiveness of this solution.

[0049] Specifically, in this test example, the mapping relationship between subjective evaluations and objective indicators (i.e., influencing factors) during the coasting braking process of seven new energy vehicles under different energy recovery modes was compared. The subjects of the subjective evaluation were six professional test engineers, and the evaluation data of each professional test engineer was collected as the subjective evaluation sample data. The coasting energy recovery drivability test conditions were set as shown in Table 2.

[0050] Table 2. Coasting Energy Recovery Drivability Test Conditions

[0051]

[0052] Taking the tests under the maximum energy recovery intensity mode of different tested vehicle models as an example, the test results of the four influencing factors (i.e., objective indicators) are shown in Tables 3, 4, 5 and 6 below.

[0053] Table 3. Real-world test results of maximum deceleration under maximum energy recovery mode.

[0054]

[0055] Table 4. Real-vehicle test results of maximum deceleration rate under maximum energy recovery mode.

[0056]

[0057] Table 5. Real-world test results of deceleration coasting distance under maximum energy recovery mode.

[0058]

[0059] Table 6. Real-vehicle test results of steady-state percentage of deceleration rate under maximum energy recovery mode.

[0060]

[0061] A comparative analysis was conducted between the output of the subjective evaluation prediction model and the measured subjective scores. The subjective scores included 1296 sample data points. The following table lists a comparison between the prediction model results and the average actual subjective scores for different initial velocities under the maximum energy recovery intensity mode, as shown in the appendix. Figure 2 As shown, the vertical distance between the output curve of the prediction model and the distribution curve of subjective ratings is used as the prediction accuracy; the smaller the vertical distance, the higher the prediction accuracy. Overall, the prediction model results are consistent with the trend of actual subjective ratings, and the prediction accuracy is good.

[0062] This embodiment also provides a drivability prediction model for electric vehicle coasting braking energy recovery, which is modeled using the modeling method for an electric vehicle coasting braking energy recovery drivability prediction model as described above; it includes a linearized model: ;in, , … These are the maximum decelerations. The linear regression parameters and the rate of change of maximum deceleration The linear regression parameters and deceleration coasting distance The linear regression parameters and maximum deceleration Quadratic regression parameters, maximum rate of change of deceleration Quadratic regression parameters, deceleration coasting distance Quadratic regression parameters, steady-state proportion of rate of change of deceleration The regression parameters, This is a disturbance term.

[0063] This embodiment provides a drivability prediction model and modeling method for electric vehicle coasting braking energy recovery, which can comprehensively and accurately evaluate and characterize the drivability of the coasting braking energy recovery system, with high evaluation consistency. Specifically, firstly, this scheme analyzes the coasting deceleration characteristics of the electric vehicle coasting energy recovery system and proposes characteristic indicators (influence factors). It then uses the correlation coefficient method to screen out influence factors with high correlation to subjective evaluation results and quantifies them. The influence factors specifically designed in this scheme can specifically characterize the coasting deceleration characteristics of the coasting energy recovery system and provide an intuitive and accurate digital expression of these characteristics. Furthermore, the screened influence factors are correlated with the subjective evaluation results, effectively maintaining high evaluation consistency and further realizing the evaluation of the drivability of the electric vehicle coasting energy recovery system.

[0064] Second, this scheme establishes a mapping relationship between quantified influencing factors and non-quantified subjective results, providing a basis for parameter calibration and optimization of deceleration control in electric vehicle coasting energy recovery systems from a human factors engineering perspective. Third, the explanatory variables (influencing factors) selected in this scheme and the regression parameters determined through nonlinear regression analysis can explain more than 77% of the data, and the model has a good fit.

[0065] The above descriptions are merely embodiments of the present invention. Commonly known structures and characteristics of the solutions are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, under the guidance of this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent.

Claims

1. A modeling method for predicting the drivability of an electric vehicle's coasting braking energy recovery model, characterized in that, Includes the following steps: Step 1: Analyze the coasting braking process of the electric vehicle coasting energy recovery system and extract influencing factors and subjective evaluation sample data; the influencing factors are characteristic indicators related to the coasting deceleration characteristics of the electric vehicle coasting energy recovery system. The subjective evaluation sample data includes deceleration intensity evaluation data, deceleration response evaluation data, and deceleration impact evaluation data; the influencing factors include the maximum deceleration corresponding to the deceleration intensity evaluation data. Extreme deceleration Root mean square of deceleration The time required to reach the deceleration inflection point and the deceleration coasting distance corresponding to the deceleration response evaluation data; the maximum rate of change of deceleration corresponding to the deceleration impact evaluation data. Range of deceleration rate of change Root mean square of the rate of change of deceleration Steady-state proportion of the rate of change of deceleration ; Step 2: Quantify the impact factors and, in conjunction with subjective evaluation sample data, screen the impact factors. When screening influencing factors, correlation analysis is used; the correlation analysis includes the following sub-steps: S1: Determine the correlation between the influencing factors and the subjective evaluation sample data through subjective judgment; S2: Visualize the quantified impact factor data using graph correlation analysis and determine the correlation between the data; S3: If the correlation in S2 is linear, then perform correlation analysis directly; if the correlation in S2 is curvilinear, then use linear regression to transform the nonlinear variable relationship into a linear variable relationship, and then perform linear correlation analysis. Step 3: Establish a subjective evaluation prediction model and use the selected influencing factors as explanatory variables for the model; optimize the model using these explanatory variables to obtain the final subjective evaluation prediction model; the established subjective evaluation prediction model is as follows: ;in, , … These are the maximum decelerations. The linear regression parameters and the rate of change of maximum deceleration The linear regression parameters and deceleration coasting distance The linear regression parameters and maximum deceleration Quadratic regression parameters, maximum rate of change of deceleration Quadratic regression parameters, deceleration coasting distance Quadratic regression parameters, steady-state proportion of rate of change of deceleration The regression parameters, This is a disturbance term.

2. The modeling method for the electric vehicle coasting braking energy recovery drivability prediction model according to claim 1, characterized in that, It also includes step 4: inputting the test data into the subjective evaluation prediction model and evaluating the subjective evaluation prediction model.

3. The modeling method for the electric vehicle coasting braking energy recovery drivability prediction model according to claim 1, characterized in that, When quantifying impact factors, the following formula should be used: Where j represents the impact intensity, in units of... ; 'a' represents vehicle acceleration, in units of... V represents vehicle speed, in units of... t represents time, in seconds (s). ; ;in, For the deceleration range, unit ; , These are the maximum and minimum deceleration values, in units of... ; For the range of the rate of change of deceleration, unit ; , These are the maximum and minimum values ​​of the rate of change of deceleration, in units of... ; ;in, For the root mean square of the deceleration, in units ; , These are the start and end times, in units. ; ;in, The percentage of the steady-state rate of change of deceleration is dimensionless. The moment when the rate of change of deceleration first reaches a steady state, unit ; For the steady-state end of the rate of change of deceleration, the unit ; Deceleration driving distance The distance traveled by an electric vehicle's coasting energy recovery system from the initial deceleration state of coasting braking to the first steady state; Time required to reach the deceleration inflection point ;in, The moment when the initial state of coasting braking begins in the electric vehicle coasting energy recovery system; This refers to the moment when the deceleration reaches its inflection point during coasting braking.

4. The modeling method for the electric vehicle coasting braking energy recovery drivability prediction model according to claim 1, characterized in that, The selection criteria for the explanatory variables include: selecting the correlation coefficient from the correlation analysis method. The influencing factors were used as explanatory variables.

5. The modeling method for the electric vehicle coasting braking energy recovery drivability prediction model according to claim 1, characterized in that, In step 3, the final subjective evaluation prediction model is: 。

Images